Research On Data Compression Of Power Quality Disturbances Based On Compressed Sensing

The Monitoring System Power quality monitoring system is facing a huge challenge, massive monitoring data that occupies a lot of storage space and enhances the requirement to the remote data transmission. For saving the storage space and reducing the duration of remote transmission, this paper studies the compressed sensing theory deeply, on this basis, proposing a new data compression method based on compressed sensing theory, applying it to the power quality data compression, getting a good compression effect.In this thesis, analyzing the data compression methods in detail at home and abroad, studying the compressed sensing theory, including the determination of the signal sparse representation, measurement matrix and sensing matrix and reconstruction algorithms. Choosing the orthogonal matching pursuit algorithm as the reconstruction algorithm in compressed sensing. In order to find the optimal sparse degree, this paper proposes a sparse degree adaptive orthogonal matching pursuit algorithm, this algorithm can automatically find the optimal sparse degree to improve the quality of reconstruction and shorten the reconstruction time.Proposing a new data compression method based on compressed sensing theory. Using Fourier transform matrix as a sparse matrix, gauss random matrix as a measurement matrix, orthogonal matching pursuit(OMP) algorithm as a reconstruction algorithm, to implement the compressed sensing theory for power quality disturbance signal data compression. Compressed sensing can complete the sampling and compression at the same time, avoiding the sampling resources waste as a result. Even under the Nyquist sampling frequency, compressed sensing theory can also recover original signal.Combining wavelet and compression sensing theory, getting the data compression method of compressed sensing based on wavelet de-noising. In thecase of noise, using wavelet de-noising to preprocess the signal at first, then compressing the signal with compressed sensing theory. Simulation results show that compressed sensing data compression method based on wavelet de-noising improves the compression ratio, increases the signal-to-noise ratio of the signal reconstruction and energy recovery, reduces the mean square error.Reconstruction effect is better.